25:02:22 23:01:24

.env.example

@@ -10,9 +10,9 @@
 
 # 数据库相关配置
 # 数据库连接字符串
-# SQL_DSN=mysql://user:password@tcp(127.0.0.1:3306)/dbname?parseTime=true
+# SQL_DSN=user:password@tcp(127.0.0.1:3306)/dbname?parseTime=true
 # 日志数据库连接字符串
-# LOG_SQL_DSN=mysql://user:password@tcp(127.0.0.1:3306)/logdb?parseTime=true
+# LOG_SQL_DSN=user:password@tcp(127.0.0.1:3306)/logdb?parseTime=true
 # SQLite数据库路径
 # SQLITE_PATH=/path/to/sqlite.db
 # 数据库最大空闲连接数

README.en.md

@@ -89,6 +89,8 @@ You can add custom models gpt-4-gizmo-* in channels. These are third-party model
 - `MAX_FILE_DOWNLOAD_MB`: Maximum file download size in MB, default `20`
 - `CRYPTO_SECRET`: Encryption key for encrypting database content
 - `AZURE_DEFAULT_API_VERSION`: Azure channel default API version, if not specified in channel settings, use this version, default `2024-12-01-preview`
+- `NOTIFICATION_LIMIT_DURATION_MINUTE`: Duration of notification limit in minutes, default `10`
+- `NOTIFY_LIMIT_COUNT`: Maximum number of user notifications in the specified duration, default `2`
 
 ## Deployment
 

Rerank.md

@@ -13,7 +13,7 @@ Request:
 
 ```json
 {
-  "model": "rerank-multilingual-v3.0",
+  "model": "jina-reranker-v2-base-multilingual",
   "query": "What is the capital of the United States?",
   "top_n": 3,
   "documents": [

VERSION

@@ -1 +0,0 @@
-v0.4.7.2.1

common/constants.go

@@ -101,7 +101,7 @@ var PreConsumedQuota = 500
 
 var RetryTimes = 0
 
-var RootUserEmail = ""
+//var RootUserEmail = ""
 
 var IsMasterNode = os.Getenv("NODE_TYPE") != "slave"
 

common/go-channel.go

@@ -1,22 +1,9 @@
 package common
 
 import (
-	"fmt"
-	"runtime/debug"
 	"time"
 )
 
-func SafeGoroutine(f func()) {
-	go func() {
-		defer func() {
-			if r := recover(); r != nil {
-				SysError(fmt.Sprintf("child goroutine panic occured: error: %v, stack: %s", r, string(debug.Stack())))
-			}
-		}()
-		f()
-	}()
-}
-
 func SafeSendBool(ch chan bool, value bool) (closed bool) {
 	defer func() {
 		// Recover from panic if one occured. A panic would mean the channel was closed.

common/logger.go

@@ -4,6 +4,7 @@ import (
 	"context"
 	"encoding/json"
 	"fmt"
+	"github.com/bytedance/gopkg/util/gopool"
 	"github.com/gin-gonic/gin"
 	"io"
 	"log"
@@ -80,9 +81,9 @@ func logHelper(ctx context.Context, level string, msg string) {
 	if logCount > maxLogCount && !setupLogWorking {
 		logCount = 0
 		setupLogWorking = true
-		go func() {
+		gopool.Go(func() {
 			SetupLogger()
-		}()
+		})
 	}
 }
 
@@ -100,6 +101,14 @@ func LogQuota(quota int) string {
 	}
 }
 
+func FormatQuota(quota int) string {
+	if DisplayInCurrencyEnabled {
+		return fmt.Sprintf("＄%.6f", float64(quota)/QuotaPerUnit)
+	} else {
+		return fmt.Sprintf("%d", quota)
+	}
+}
+
 // LogJson 仅供测试使用 only for test
 func LogJson(ctx context.Context, msg string, obj any) {
 	jsonStr, err := json.Marshal(obj)

common/model-ratio.go

@@ -233,7 +233,11 @@ var (
 	modelRatioMapMutex                    = sync.RWMutex{}
 )
 
-var CompletionRatio map[string]float64 = nil
+var (
+	CompletionRatio      map[string]float64 = nil
+	CompletionRatioMutex                    = sync.RWMutex{}
+)
+
 var defaultCompletionRatio = map[string]float64{
 	"gpt-4-gizmo-*":  2,
 	"gpt-4o-gizmo-*": 3,
@@ -334,10 +338,17 @@ func GetDefaultModelRatioMap() map[string]float64 {
 	return defaultModelRatio
 }
 
-func CompletionRatio2JSONString() string {
+func GetCompletionRatioMap() map[string]float64 {
+	CompletionRatioMutex.Lock()
+	defer CompletionRatioMutex.Unlock()
 	if CompletionRatio == nil {
 		CompletionRatio = defaultCompletionRatio
 	}
+	return CompletionRatio
+}
+
+func CompletionRatio2JSONString() string {
+	GetCompletionRatioMap()
 	jsonBytes, err := json.Marshal(CompletionRatio)
 	if err != nil {
 		SysError("error marshalling completion ratio: " + err.Error())
@@ -346,11 +357,15 @@ func CompletionRatio2JSONString() string {
 }
 
 func UpdateCompletionRatioByJSONString(jsonStr string) error {
+	CompletionRatioMutex.Lock()
+	defer CompletionRatioMutex.Unlock()
 	CompletionRatio = make(map[string]float64)
 	return json.Unmarshal([]byte(jsonStr), &CompletionRatio)
 }
 
 func GetCompletionRatio(name string) float64 {
+	GetCompletionRatioMap()
+
 	if strings.Contains(name, "/") {
 		if ratio, ok := CompletionRatio[name]; ok {
 			return ratio
@@ -476,24 +491,3 @@ func GetAudioCompletionRatio(name string) float64 {
 	}
 	return 2
 }
-
-//func GetAudioPricePerMinute(name string) float64 {
-//	if strings.HasPrefix(name, "gpt-4o-realtime") {
-//		return 0.06
-//	}
-//	return 0.06
-//}
-//
-//func GetAudioCompletionPricePerMinute(name string) float64 {
-//	if strings.HasPrefix(name, "gpt-4o-realtime") {
-//		return 0.24
-//	}
-//	return 0.24
-//}
-
-func GetCompletionRatioMap() map[string]float64 {
-	if CompletionRatio == nil {
-		CompletionRatio = defaultCompletionRatio
-	}
-	return CompletionRatio
-}

constant/env.go

@@ -29,6 +29,9 @@ var GeminiModelMap = map[string]string{
 
 var GeminiVisionMaxImageNum = common.GetEnvOrDefault("GEMINI_VISION_MAX_IMAGE_NUM", 16)
 
+var NotifyLimitCount = common.GetEnvOrDefault("NOTIFY_LIMIT_COUNT", 2)
+var NotificationLimitDurationMinute = common.GetEnvOrDefault("NOTIFICATION_LIMIT_DURATION_MINUTE", 10)
+
 func InitEnv() {
 	modelVersionMapStr := strings.TrimSpace(os.Getenv("GEMINI_MODEL_MAP"))
 	if modelVersionMapStr == "" {
@@ -44,5 +47,5 @@ func InitEnv() {
 	}
 }
 
-// 是否生成初始令牌，默认关闭。
+// GenerateDefaultToken 是否生成初始令牌，默认关闭。
 var GenerateDefaultToken = common.GetEnvOrDefaultBool("GENERATE_DEFAULT_TOKEN", false)

constant/user_setting.go

@@ -0,0 +1,14 @@
+package constant
+
+var (
+	UserSettingNotifyType            = "notify_type"             // QuotaWarningType 额度预警类型
+	UserSettingQuotaWarningThreshold = "quota_warning_threshold" // QuotaWarningThreshold 额度预警阈值
+	UserSettingWebhookUrl            = "webhook_url"             // WebhookUrl webhook地址
+	UserSettingWebhookSecret         = "webhook_secret"          // WebhookSecret webhook密钥
+	UserSettingNotificationEmail     = "notification_email"      // NotificationEmail 通知邮箱地址
+)
+
+var (
+	NotifyTypeEmail   = "email"   // Email 邮件
+	NotifyTypeWebhook = "webhook" // Webhook
+)

controller/channel-test.go

@@ -238,9 +238,7 @@ var testAllChannelsLock sync.Mutex
 var testAllChannelsRunning bool = false
 
 func testAllChannels(notify bool) error {
-	if common.RootUserEmail == "" {
-		common.RootUserEmail = model.GetRootUserEmail()
-	}
+
 	testAllChannelsLock.Lock()
 	if testAllChannelsRunning {
 		testAllChannelsLock.Unlock()
@@ -295,10 +293,7 @@ func testAllChannels(notify bool) error {
 		testAllChannelsRunning = false
 		testAllChannelsLock.Unlock()
 		if notify {
-			err := common.SendEmail("通道测试完成", common.RootUserEmail, "通道测试完成，如果没有收到禁用通知，说明所有通道都正常")
-			if err != nil {
-				common.SysError(fmt.Sprintf("failed to send email: %s", err.Error()))
-			}
+			service.NotifyRootUser(dto.NotifyTypeChannelTest, "通道测试完成", "所有通道测试已完成")
 		}
 	})
 	return nil

controller/pricing.go

@@ -17,7 +17,7 @@ func GetPricing(c *gin.Context) {
 	}
 	var group string
 	if exists {
-		user, err := model.GetUserById(userId.(int), false)
+		user, err := model.GetUserCache(userId.(int))
 		if err == nil {
 			group = user.Group
 		}

controller/relay.go

@@ -24,7 +24,7 @@ func relayHandler(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode
 	var err *dto.OpenAIErrorWithStatusCode
 	switch relayMode {
 	case relayconstant.RelayModeImagesGenerations:
-		err = relay.ImageHelper(c, relayMode)
+		err = relay.ImageHelper(c)
 	case relayconstant.RelayModeAudioSpeech:
 		fallthrough
 	case relayconstant.RelayModeAudioTranslation:

controller/user.go

@@ -4,6 +4,7 @@ import (
 	"encoding/json"
 	"fmt"
 	"net/http"
+	"net/url"
 	"one-api/common"
 	"one-api/model"
 	"one-api/setting"
@@ -471,7 +472,7 @@ func GetUserModels(c *gin.Context) {
 	if err != nil {
 		id = c.GetInt("id")
 	}
-	user, err := model.GetUserById(id, true)
+	user, err := model.GetUserCache(id)
 	if err != nil {
 		c.JSON(http.StatusOK, gin.H{
 			"success": false,
@@ -869,9 +870,6 @@ func EmailBind(c *gin.Context) {
 		})
 		return
 	}
-	if user.Role == common.RoleRootUser {
-		common.RootUserEmail = email
-	}
 	c.JSON(http.StatusOK, gin.H{
 		"success": true,
 		"message": "",
@@ -913,3 +911,115 @@ func TopUp(c *gin.Context) {
 	})
 	return
 }
+
+type UpdateUserSettingRequest struct {
+	QuotaWarningType      string  `json:"notify_type"`
+	QuotaWarningThreshold float64 `json:"quota_warning_threshold"`
+	WebhookUrl            string  `json:"webhook_url,omitempty"`
+	WebhookSecret         string  `json:"webhook_secret,omitempty"`
+	NotificationEmail     string  `json:"notification_email,omitempty"`
+}
+
+func UpdateUserSetting(c *gin.Context) {
+	var req UpdateUserSettingRequest
+	if err := c.ShouldBindJSON(&req); err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "无效的参数",
+		})
+		return
+	}
+
+	// 验证预警类型
+	if req.QuotaWarningType != constant.NotifyTypeEmail && req.QuotaWarningType != constant.NotifyTypeWebhook {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "无效的预警类型",
+		})
+		return
+	}
+
+	// 验证预警阈值
+	if req.QuotaWarningThreshold <= 0 {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "预警阈值必须大于0",
+		})
+		return
+	}
+
+	// 如果是webhook类型,验证webhook地址
+	if req.QuotaWarningType == constant.NotifyTypeWebhook {
+		if req.WebhookUrl == "" {
+			c.JSON(http.StatusOK, gin.H{
+				"success": false,
+				"message": "Webhook地址不能为空",
+			})
+			return
+		}
+		// 验证URL格式
+		if _, err := url.ParseRequestURI(req.WebhookUrl); err != nil {
+			c.JSON(http.StatusOK, gin.H{
+				"success": false,
+				"message": "无效的Webhook地址",
+			})
+			return
+		}
+	}
+
+	// 如果是邮件类型，验证邮箱地址
+	if req.QuotaWarningType == constant.NotifyTypeEmail && req.NotificationEmail != "" {
+		// 验证邮箱格式
+		if !strings.Contains(req.NotificationEmail, "@") {
+			c.JSON(http.StatusOK, gin.H{
+				"success": false,
+				"message": "无效的邮箱地址",
+			})
+			return
+		}
+	}
+
+	userId := c.GetInt("id")
+	user, err := model.GetUserById(userId, true)
+	if err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": err.Error(),
+		})
+		return
+	}
+
+	// 构建设置
+	settings := map[string]interface{}{
+		constant.UserSettingNotifyType:            req.QuotaWarningType,
+		constant.UserSettingQuotaWarningThreshold: req.QuotaWarningThreshold,
+	}
+
+	// 如果是webhook类型,添加webhook相关设置
+	if req.QuotaWarningType == constant.NotifyTypeWebhook {
+		settings[constant.UserSettingWebhookUrl] = req.WebhookUrl
+		if req.WebhookSecret != "" {
+			settings[constant.UserSettingWebhookSecret] = req.WebhookSecret
+		}
+	}
+
+	// 如果提供了通知邮箱，添加到设置中
+	if req.QuotaWarningType == constant.NotifyTypeEmail && req.NotificationEmail != "" {
+		settings[constant.UserSettingNotificationEmail] = req.NotificationEmail
+	}
+
+	// 更新用户设置
+	user.SetSetting(settings)
+	if err := user.Update(false); err != nil {
+		c.JSON(http.StatusOK, gin.H{
+			"success": false,
+			"message": "更新设置失败: " + err.Error(),
+		})
+		return
+	}
+
+	c.JSON(http.StatusOK, gin.H{
+		"success": true,
+		"message": "设置已更新",
+	})
+}

docker-compose.yml

@@ -24,7 +24,7 @@ services:
       - redis
       - mysql
     healthcheck:
-      test: [ "CMD-SHELL", "wget -q -O - http://localhost:3000/api/status | grep -o '\"success\":\\s*true' | awk -F: '{print $2}'" ]
+      test: ["CMD-SHELL", "wget -q -O - http://localhost:3000/api/status | grep -o '\"success\":\\s*true' | awk -F: '{print $$2}'"]
       interval: 30s
       timeout: 10s
       retries: 3

dto/notify.go

@@ -0,0 +1,25 @@
+package dto
+
+type Notify struct {
+	Type    string        `json:"type"`
+	Title   string        `json:"title"`
+	Content string        `json:"content"`
+	Values  []interface{} `json:"values"`
+}
+
+const ContentValueParam = "{{value}}"
+
+const (
+	NotifyTypeQuotaExceed   = "quota_exceed"
+	NotifyTypeChannelUpdate = "channel_update"
+	NotifyTypeChannelTest   = "channel_test"
+)
+
+func NewNotify(t string, title string, content string, values []interface{}) Notify {
+	return Notify{
+		Type:    t,
+		Title:   title,
+		Content: content,
+		Values:  values,
+	}
+}

dto/openai_request.go

@@ -18,6 +18,8 @@ type GeneralOpenAIRequest struct {
 	Model               string          `json:"model,omitempty"`
 	Messages            []Message       `json:"messages,omitempty"`
 	Prompt              any             `json:"prompt,omitempty"`
+	Prefix              any             `json:"prefix,omitempty"`
+	Suffix              any             `json:"suffix,omitempty"`
 	Stream              bool            `json:"stream,omitempty"`
 	StreamOptions       *StreamOptions  `json:"stream_options,omitempty"`
 	MaxTokens           uint            `json:"max_tokens,omitempty"`
@@ -86,18 +88,20 @@ func (r GeneralOpenAIRequest) ParseInput() []string {
 }
 
 type Message struct {
-	Role             string          `json:"role"`
-	Content          json.RawMessage `json:"content"`
-	Name             *string         `json:"name,omitempty"`
-	Prefix           *bool           `json:"prefix,omitempty"`
-	ReasoningContent string          `json:"reasoning_content,omitempty"`
-	ToolCalls        json.RawMessage `json:"tool_calls,omitempty"`
-	ToolCallId       string          `json:"tool_call_id,omitempty"`
+	Role                string          `json:"role"`
+	Content             json.RawMessage `json:"content"`
+	Name                *string         `json:"name,omitempty"`
+	Prefix              *bool           `json:"prefix,omitempty"`
+	ReasoningContent    string          `json:"reasoning_content,omitempty"`
+	ToolCalls           json.RawMessage `json:"tool_calls,omitempty"`
+	ToolCallId          string          `json:"tool_call_id,omitempty"`
+	parsedContent       []MediaContent
+	parsedStringContent *string
 }
 
 type MediaContent struct {
 	Type       string `json:"type"`
-	Text       string `json:"text"`
+	Text       string `json:"text,omitempty"`
 	ImageUrl   any    `json:"image_url,omitempty"`
 	InputAudio any    `json:"input_audio,omitempty"`
 }
@@ -146,6 +150,9 @@ func (m *Message) SetToolCalls(toolCalls any) {
 }
 
 func (m *Message) StringContent() string {
+	if m.parsedStringContent != nil {
+		return *m.parsedStringContent
+	}
 	var stringContent string
 	if err := json.Unmarshal(m.Content, &stringContent); err == nil {
 		return stringContent
@@ -156,78 +163,113 @@ func (m *Message) StringContent() string {
 func (m *Message) SetStringContent(content string) {
 	jsonContent, _ := json.Marshal(content)
 	m.Content = jsonContent
+	m.parsedStringContent = &content
+	m.parsedContent = nil
+}
+
+func (m *Message) SetMediaContent(content []MediaContent) {
+	jsonContent, _ := json.Marshal(content)
+	m.Content = jsonContent
+	m.parsedContent = nil
+	m.parsedStringContent = nil
 }
 
 func (m *Message) IsStringContent() bool {
+	if m.parsedStringContent != nil {
+		return true
+	}
 	var stringContent string
 	if err := json.Unmarshal(m.Content, &stringContent); err == nil {
+		m.parsedStringContent = &stringContent
 		return true
 	}
 	return false
 }
 
 func (m *Message) ParseContent() []MediaContent {
+	if m.parsedContent != nil {
+		return m.parsedContent
+	}
+
 	var contentList []MediaContent
+
+	// 先尝试解析为字符串
 	var stringContent string
 	if err := json.Unmarshal(m.Content, &stringContent); err == nil {
-		contentList = append(contentList, MediaContent{
+		contentList = []MediaContent{{
 			Type: ContentTypeText,
 			Text: stringContent,
-		})
+		}}
+		m.parsedContent = contentList
 		return contentList
 	}
-	var arrayContent []json.RawMessage
+
+	// 尝试解析为数组
+	var arrayContent []map[string]interface{}
 	if err := json.Unmarshal(m.Content, &arrayContent); err == nil {
 		for _, contentItem := range arrayContent {
-			var contentMap map[string]any
-			if err := json.Unmarshal(contentItem, &contentMap); err != nil {
+			contentType, ok := contentItem["type"].(string)
+			if !ok {
 				continue
 			}
-			switch contentMap["type"] {
+
+			switch contentType {
 			case ContentTypeText:
-				if subStr, ok := contentMap["text"].(string); ok {
+				if text, ok := contentItem["text"].(string); ok {
 					contentList = append(contentList, MediaContent{
 						Type: ContentTypeText,
-						Text: subStr,
+						Text: text,
 					})
 				}
+
 			case ContentTypeImageURL:
-				if subObj, ok := contentMap["image_url"].(map[string]any); ok {
-					detail, ok := subObj["detail"]
-					if ok {
-						subObj["detail"] = detail.(string)
-					} else {
-						subObj["detail"] = "high"
-					}
+				imageUrl := contentItem["image_url"]
+				switch v := imageUrl.(type) {
+				case string:
 					contentList = append(contentList, MediaContent{
 						Type: ContentTypeImageURL,
 						ImageUrl: MessageImageUrl{
-							Url:    subObj["url"].(string),
-							Detail: subObj["detail"].(string),
-						},
-					})
-				} else if url, ok := contentMap["image_url"].(string); ok {
-					contentList = append(contentList, MediaContent{
-						Type: ContentTypeImageURL,
-						ImageUrl: MessageImageUrl{
-							Url:    url,
+							Url:    v,
 							Detail: "high",
 						},
 					})
+				case map[string]interface{}:
+					url, ok1 := v["url"].(string)
+					detail, ok2 := v["detail"].(string)
+					if !ok2 {
+						detail = "high"
+					}
+					if ok1 {
+						contentList = append(contentList, MediaContent{
+							Type: ContentTypeImageURL,
+							ImageUrl: MessageImageUrl{
+								Url:    url,
+								Detail: detail,
+							},
+						})
+					}
 				}
+
 			case ContentTypeInputAudio:
-				if subObj, ok := contentMap["input_audio"].(map[string]any); ok {
-					contentList = append(contentList, MediaContent{
-						Type: ContentTypeInputAudio,
-						InputAudio: MessageInputAudio{
-							Data:   subObj["data"].(string),
-							Format: subObj["format"].(string),
-						},
-					})
+				if audioData, ok := contentItem["input_audio"].(map[string]interface{}); ok {
+					data, ok1 := audioData["data"].(string)
+					format, ok2 := audioData["format"].(string)
+					if ok1 && ok2 {
+						contentList = append(contentList, MediaContent{
+							Type: ContentTypeInputAudio,
+							InputAudio: MessageInputAudio{
+								Data:   data,
+								Format: format,
+							},
+						})
+					}
 				}
 			}
 		}
-		return contentList
 	}
-	return nil
+
+	if len(contentList) > 0 {
+		m.parsedContent = contentList
+	}
+	return contentList
 }

dto/openai_response.go

@@ -62,9 +62,10 @@ type ChatCompletionsStreamResponseChoice struct {
 }
 
 type ChatCompletionsStreamResponseChoiceDelta struct {
-	Content   *string    `json:"content,omitempty"`
-	Role      string     `json:"role,omitempty"`
-	ToolCalls []ToolCall `json:"tool_calls,omitempty"`
+	Content          *string    `json:"content,omitempty"`
+	ReasoningContent *string    `json:"reasoning_content,omitempty"`
+	Role             string     `json:"role,omitempty"`
+	ToolCalls        []ToolCall `json:"tool_calls,omitempty"`
 }
 
 func (c *ChatCompletionsStreamResponseChoiceDelta) SetContentString(s string) {
@@ -78,6 +79,13 @@ func (c *ChatCompletionsStreamResponseChoiceDelta) GetContentString() string {
 	return *c.Content
 }
 
+func (c *ChatCompletionsStreamResponseChoiceDelta) GetReasoningContent() string {
+	if c.ReasoningContent == nil {
+		return ""
+	}
+	return *c.ReasoningContent
+}
+
 type ToolCall struct {
 	// Index is not nil only in chat completion chunk object
 	Index    *int         `json:"index,omitempty"`

main.go

@@ -119,9 +119,9 @@ func main() {
 	}
 
 	if os.Getenv("ENABLE_PPROF") == "true" {
-		go func() {
+		gopool.Go(func() {
 			log.Println(http.ListenAndServe("0.0.0.0:8005", nil))
-		}()
+		})
 		go common.Monitor()
 		common.SysLog("pprof enabled")
 	}

middleware/distributor.go

@@ -135,17 +135,14 @@ func getModelRequest(c *gin.Context) (*ModelRequest, bool, error) {
 			midjourneyRequest := dto.MidjourneyRequest{}
 			err = common.UnmarshalBodyReusable(c, &midjourneyRequest)
 			if err != nil {
-				abortWithMidjourneyMessage(c, http.StatusBadRequest, constant.MjErrorUnknown, "无效的请求, "+err.Error())
 				return nil, false, err
 			}
 			midjourneyModel, mjErr, success := service.GetMjRequestModel(relayMode, &midjourneyRequest)
 			if mjErr != nil {
-				abortWithMidjourneyMessage(c, http.StatusBadRequest, mjErr.Code, mjErr.Description)
 				return nil, false, fmt.Errorf(mjErr.Description)
 			}
 			if midjourneyModel == "" {
 				if !success {
-					abortWithMidjourneyMessage(c, http.StatusBadRequest, constant.MjErrorUnknown, "无效的请求, 无法解析模型")
 					return nil, false, fmt.Errorf("无效的请求, 无法解析模型")
 				} else {
 					// task fetch, task fetch by condition, notify
@@ -170,7 +167,6 @@ func getModelRequest(c *gin.Context) (*ModelRequest, bool, error) {
 		err = common.UnmarshalBodyReusable(c, &modelRequest)
 	}
 	if err != nil {
-		abortWithOpenAiMessage(c, http.StatusBadRequest, "无效的请求, "+err.Error())
 		return nil, false, errors.New("无效的请求, " + err.Error())
 	}
 	if strings.HasPrefix(c.Request.URL.Path, "/v1/realtime") {

model/option.go

@@ -84,7 +84,7 @@ func InitOptionMap() {
 	common.OptionMap["QuotaForInviter"] = strconv.Itoa(common.QuotaForInviter)
 	common.OptionMap["QuotaForInvitee"] = strconv.Itoa(common.QuotaForInvitee)
 	common.OptionMap["QuotaRemindThreshold"] = strconv.Itoa(common.QuotaRemindThreshold)
-	common.OptionMap["PreConsumedQuota"] = strconv.Itoa(common.PreConsumedQuota)
+	common.OptionMap["ShouldPreConsumedQuota"] = strconv.Itoa(common.PreConsumedQuota)
 	common.OptionMap["ModelRatio"] = common.ModelRatio2JSONString()
 	common.OptionMap["ModelPrice"] = common.ModelPrice2JSONString()
 	common.OptionMap["GroupRatio"] = setting.GroupRatio2JSONString()
@@ -306,7 +306,7 @@ func updateOptionMap(key string, value string) (err error) {
 		common.QuotaForInvitee, _ = strconv.Atoi(value)
 	case "QuotaRemindThreshold":
 		common.QuotaRemindThreshold, _ = strconv.Atoi(value)
-	case "PreConsumedQuota":
+	case "ShouldPreConsumedQuota":
 		common.PreConsumedQuota, _ = strconv.Atoi(value)
 	case "RetryTimes":
 		common.RetryTimes, _ = strconv.Atoi(value)

model/token.go

@@ -3,13 +3,11 @@ package model
 import (
 	"errors"
 	"fmt"
-	"github.com/bytedance/gopkg/util/gopool"
-	"gorm.io/gorm"
 	"one-api/common"
-	relaycommon "one-api/relay/common"
-	"one-api/setting"
-	"strconv"
 	"strings"
+
+	"github.com/bytedance/gopkg/util/gopool"
+	"gorm.io/gorm"
 )
 
 type Token struct {
@@ -322,80 +320,3 @@ func decreaseTokenQuota(id int, quota int) (err error) {
 	).Error
 	return err
 }
-
-func PreConsumeTokenQuota(relayInfo *relaycommon.RelayInfo, quota int) error {
-	if quota < 0 {
-		return errors.New("quota 不能为负数！")
-	}
-	if relayInfo.IsPlayground {
-		return nil
-	}
-	//if relayInfo.TokenUnlimited {
-	//	return nil
-	//}
-	token, err := GetTokenById(relayInfo.TokenId)
-	if err != nil {
-		return err
-	}
-	if !relayInfo.TokenUnlimited && token.RemainQuota < quota {
-		return errors.New("令牌额度不足")
-	}
-	err = DecreaseTokenQuota(relayInfo.TokenId, relayInfo.TokenKey, quota)
-	if err != nil {
-		return err
-	}
-	return nil
-}
-
-func PostConsumeQuota(relayInfo *relaycommon.RelayInfo, userQuota int, quota int, preConsumedQuota int, sendEmail bool) (err error) {
-
-	if quota > 0 {
-		err = DecreaseUserQuota(relayInfo.UserId, quota)
-	} else {
-		err = IncreaseUserQuota(relayInfo.UserId, -quota)
-	}
-	if err != nil {
-		return err
-	}
-
-	if !relayInfo.IsPlayground {
-		if quota > 0 {
-			err = DecreaseTokenQuota(relayInfo.TokenId, relayInfo.TokenKey, quota)
-		} else {
-			err = IncreaseTokenQuota(relayInfo.TokenId, relayInfo.TokenKey, -quota)
-		}
-		if err != nil {
-			return err
-		}
-	}
-
-	if sendEmail {
-		if (quota + preConsumedQuota) != 0 {
-			quotaTooLow := userQuota >= common.QuotaRemindThreshold && userQuota-(quota+preConsumedQuota) < common.QuotaRemindThreshold
-			noMoreQuota := userQuota-(quota+preConsumedQuota) <= 0
-			if quotaTooLow || noMoreQuota {
-				go func() {
-					email, err := GetUserEmail(relayInfo.UserId)
-					if err != nil {
-						common.SysError("failed to fetch user email: " + err.Error())
-					}
-					prompt := "您的额度即将用尽"
-					if noMoreQuota {
-						prompt = "您的额度已用尽"
-					}
-					if email != "" {
-						topUpLink := fmt.Sprintf("%s/topup", setting.ServerAddress)
-						err = common.SendEmail(prompt, email,
-							fmt.Sprintf("%s，当前剩余额度为 %d，为了不影响您的使用，请及时充值。<br/>充值链接：<a href='%s'>%s</a>", prompt, userQuota, topUpLink, topUpLink))
-						if err != nil {
-							common.SysError("failed to send email" + err.Error())
-						}
-						common.SysLog("user quota is low, consumed quota: " + strconv.Itoa(quota) + ", user quota: " + strconv.Itoa(userQuota))
-					}
-				}()
-			}
-		}
-	}
-
-	return nil
-}

model/token_cache.go

@@ -52,7 +52,7 @@ func cacheSetTokenField(key string, field string, value string) error {
 func cacheGetTokenByKey(key string) (*Token, error) {
 	hmacKey := common.GenerateHMAC(key)
 	if !common.RedisEnabled {
-		return nil, nil
+		return nil, fmt.Errorf("redis is not enabled")
 	}
 	var token Token
 	err := common.RedisHGetObj(fmt.Sprintf("token:%s", hmacKey), &token)

model/user.go

@@ -1,6 +1,7 @@
 package model
 
 import (
+	"encoding/json"
 	"errors"
 	"fmt"
 	"one-api/common"
@@ -38,6 +39,20 @@ type User struct {
 	InviterId        int            `json:"inviter_id" gorm:"type:int;column:inviter_id;index"`
 	DeletedAt        gorm.DeletedAt `gorm:"index"`
 	LinuxDOId        string         `json:"linux_do_id" gorm:"column:linux_do_id;index"`
+	Setting          string         `json:"setting" gorm:"type:text;column:setting"`
+}
+
+func (user *User) ToBaseUser() *UserBase {
+	cache := &UserBase{
+		Id:       user.Id,
+		Group:    user.Group,
+		Quota:    user.Quota,
+		Status:   user.Status,
+		Username: user.Username,
+		Setting:  user.Setting,
+		Email:    user.Email,
+	}
+	return cache
 }
 
 func (user *User) GetAccessToken() string {
@@ -51,6 +66,22 @@ func (user *User) SetAccessToken(token string) {
 	user.AccessToken = &token
 }
 
+func (user *User) GetSetting() map[string]interface{} {
+	if user.Setting == "" {
+		return nil
+	}
+	return common.StrToMap(user.Setting)
+}
+
+func (user *User) SetSetting(setting map[string]interface{}) {
+	settingBytes, err := json.Marshal(setting)
+	if err != nil {
+		common.SysError("failed to marshal setting: " + err.Error())
+		return
+	}
+	user.Setting = string(settingBytes)
+}
+
 // CheckUserExistOrDeleted check if user exist or deleted, if not exist, return false, nil, if deleted or exist, return true, nil
 func CheckUserExistOrDeleted(username string, email string) (bool, error) {
 	var user User
@@ -315,8 +346,8 @@ func (user *User) Update(updatePassword bool) error {
 		return err
 	}
 
-	// 更新缓存
-	return updateUserCache(user.Id, user.Username, user.Group, user.Quota, user.Status)
+	// Update cache
+	return updateUserCache(*user)
 }
 
 func (user *User) Edit(updatePassword bool) error {
@@ -344,8 +375,8 @@ func (user *User) Edit(updatePassword bool) error {
 		return err
 	}
 
-	// 更新缓存
-	return updateUserCache(user.Id, user.Username, user.Group, user.Quota, user.Status)
+	// Update cache
+	return updateUserCache(*user)
 }
 
 func (user *User) Delete() error {
@@ -371,8 +402,8 @@ func (user *User) HardDelete() error {
 // ValidateAndFill check password & user status
 func (user *User) ValidateAndFill() (err error) {
 	// When querying with struct, GORM will only query with non-zero fields,
-	// that means if your field’s value is 0, '', false or other zero values,
-	// it won’t be used to build query conditions
+	// that means if your field's value is 0, '', false or other zero values,
+	// it won't be used to build query conditions
 	password := user.Password
 	username := strings.TrimSpace(user.Username)
 	if username == "" || password == "" {
@@ -531,7 +562,6 @@ func GetUserQuota(id int, fromDB bool) (quota int, err error) {
 			return quota, nil
 		}
 		// Don't return error - fall through to DB
-		//common.SysError("failed to get user quota from cache: " + err.Error())
 	}
 	fromDB = true
 	err = DB.Model(&User{}).Where("id = ?", id).Select("quota").Find(&quota).Error
@@ -580,6 +610,35 @@ func GetUserGroup(id int, fromDB bool) (group string, err error) {
 	return group, nil
 }
 
+// GetUserSetting gets setting from Redis first, falls back to DB if needed
+func GetUserSetting(id int, fromDB bool) (settingMap map[string]interface{}, err error) {
+	var setting string
+	defer func() {
+		// Update Redis cache asynchronously on successful DB read
+		if shouldUpdateRedis(fromDB, err) {
+			gopool.Go(func() {
+				if err := updateUserSettingCache(id, setting); err != nil {
+					common.SysError("failed to update user setting cache: " + err.Error())
+				}
+			})
+		}
+	}()
+	if !fromDB && common.RedisEnabled {
+		setting, err := getUserSettingCache(id)
+		if err == nil {
+			return setting, nil
+		}
+		// Don't return error - fall through to DB
+	}
+	fromDB = true
+	err = DB.Model(&User{}).Where("id = ?", id).Select("setting").Find(&setting).Error
+	if err != nil {
+		return map[string]interface{}{}, err
+	}
+
+	return common.StrToMap(setting), nil
+}
+
 func IncreaseUserQuota(id int, quota int) (err error) {
 	if quota < 0 {
 		return errors.New("quota 不能为负数！")
@@ -641,9 +700,14 @@ func DeltaUpdateUserQuota(id int, delta int) (err error) {
 	}
 }
 
-func GetRootUserEmail() (email string) {
-	DB.Model(&User{}).Where("role = ?", common.RoleRootUser).Select("email").Find(&email)
-	return email
+//func GetRootUserEmail() (email string) {
+//	DB.Model(&User{}).Where("role = ?", common.RoleRootUser).Select("email").Find(&email)
+//	return email
+//}
+
+func GetRootUser() (user *User) {
+	DB.Where("role = ?", common.RoleRootUser).First(&user)
+	return user
 }
 
 func UpdateUserUsedQuotaAndRequestCount(id int, quota int) {
@@ -725,10 +789,10 @@ func IsLinuxDOIdAlreadyTaken(linuxDOId string) bool {
 	return !errors.Is(err, gorm.ErrRecordNotFound)
 }
 
-func (u *User) FillUserByLinuxDOId() error {
-	if u.LinuxDOId == "" {
+func (user *User) FillUserByLinuxDOId() error {
+	if user.LinuxDOId == "" {
 		return errors.New("linux do id is empty")
 	}
-	err := DB.Where("linux_do_id = ?", u.LinuxDOId).First(u).Error
+	err := DB.Where("linux_do_id = ?", user.LinuxDOId).First(user).Error
 	return err
 }

model/user_cache.go

@@ -1,206 +1,213 @@
 package model
 
 import (
+	"encoding/json"
 	"fmt"
 	"one-api/common"
 	"one-api/constant"
-	"strconv"
 	"time"
+
+	"github.com/bytedance/gopkg/util/gopool"
 )
 
-// Change UserCache struct to userCache
-type userCache struct {
+// UserBase struct remains the same as it represents the cached data structure
+type UserBase struct {
 	Id       int    `json:"id"`
 	Group    string `json:"group"`
+	Email    string `json:"email"`
 	Quota    int    `json:"quota"`
 	Status   int    `json:"status"`
-	Role     int    `json:"role"`
 	Username string `json:"username"`
+	Setting  string `json:"setting"`
 }
 
-// Rename all exported functions to private ones
-// invalidateUserCache clears all user related cache
-func invalidateUserCache(userId int) error {
-	if !common.RedisEnabled {
+func (user *UserBase) GetSetting() map[string]interface{} {
+	if user.Setting == "" {
 		return nil
 	}
-
-	keys := []string{
-		fmt.Sprintf(constant.UserGroupKeyFmt, userId),
-		fmt.Sprintf(constant.UserQuotaKeyFmt, userId),
-		fmt.Sprintf(constant.UserEnabledKeyFmt, userId),
-		fmt.Sprintf(constant.UserUsernameKeyFmt, userId),
-	}
-
-	for _, key := range keys {
-		if err := common.RedisDel(key); err != nil {
-			return fmt.Errorf("failed to delete cache key %s: %w", key, err)
-		}
-	}
-	return nil
+	return common.StrToMap(user.Setting)
 }
 
-// updateUserGroupCache updates user group cache
-func updateUserGroupCache(userId int, group string) error {
-	if !common.RedisEnabled {
-		return nil
+func (user *UserBase) SetSetting(setting map[string]interface{}) {
+	settingBytes, err := json.Marshal(setting)
+	if err != nil {
+		common.SysError("failed to marshal setting: " + err.Error())
+		return
 	}
-	return common.RedisSet(
-		fmt.Sprintf(constant.UserGroupKeyFmt, userId),
-		group,
-		time.Duration(constant.UserId2QuotaCacheSeconds)*time.Second,
-	)
+	user.Setting = string(settingBytes)
 }
 
-// updateUserQuotaCache updates user quota cache
-func updateUserQuotaCache(userId int, quota int) error {
-	if !common.RedisEnabled {
-		return nil
-	}
-	return common.RedisSet(
-		fmt.Sprintf(constant.UserQuotaKeyFmt, userId),
-		fmt.Sprintf("%d", quota),
-		time.Duration(constant.UserId2QuotaCacheSeconds)*time.Second,
-	)
+// getUserCacheKey returns the key for user cache
+func getUserCacheKey(userId int) string {
+	return fmt.Sprintf("user:%d", userId)
 }
 
-// updateUserStatusCache updates user status cache
-func updateUserStatusCache(userId int, userEnabled bool) error {
+// invalidateUserCache clears user cache
+func invalidateUserCache(userId int) error {
 	if !common.RedisEnabled {
 		return nil
 	}
-	enabled := "0"
-	if userEnabled {
-		enabled = "1"
-	}
-	return common.RedisSet(
-		fmt.Sprintf(constant.UserEnabledKeyFmt, userId),
-		enabled,
-		time.Duration(constant.UserId2StatusCacheSeconds)*time.Second,
-	)
+	return common.RedisHDelObj(getUserCacheKey(userId))
 }
 
-// updateUserNameCache updates username cache
-func updateUserNameCache(userId int, username string) error {
+// updateUserCache updates all user cache fields using hash
+func updateUserCache(user User) error {
 	if !common.RedisEnabled {
 		return nil
 	}
-	return common.RedisSet(
-		fmt.Sprintf(constant.UserUsernameKeyFmt, userId),
-		username,
+
+	return common.RedisHSetObj(
+		getUserCacheKey(user.Id),
+		user.ToBaseUser(),
 		time.Duration(constant.UserId2QuotaCacheSeconds)*time.Second,
 	)
 }
 
-// updateUserCache updates all user cache fields
-func updateUserCache(userId int, username string, userGroup string, quota int, status int) error {
-	if !common.RedisEnabled {
-		return nil
+// GetUserCache gets complete user cache from hash
+func GetUserCache(userId int) (userCache *UserBase, err error) {
+	var user *User
+	var fromDB bool
+	defer func() {
+		// Update Redis cache asynchronously on successful DB read
+		if shouldUpdateRedis(fromDB, err) && user != nil {
+			gopool.Go(func() {
+				if err := updateUserCache(*user); err != nil {
+					common.SysError("failed to update user status cache: " + err.Error())
+				}
+			})
+		}
+	}()
+
+	// Try getting from Redis first
+	userCache, err = cacheGetUserBase(userId)
+	if err == nil {
+		return userCache, nil
 	}
 
-	if err := updateUserGroupCache(userId, userGroup); err != nil {
-		return fmt.Errorf("update group cache: %w", err)
+	// If Redis fails, get from DB
+	fromDB = true
+	user, err = GetUserById(userId, false)
+	if err != nil {
+		return nil, err // Return nil and error if DB lookup fails
 	}
 
-	if err := updateUserQuotaCache(userId, quota); err != nil {
-		return fmt.Errorf("update quota cache: %w", err)
+	// Create cache object from user data
+	userCache = &UserBase{
+		Id:       user.Id,
+		Group:    user.Group,
+		Quota:    user.Quota,
+		Status:   user.Status,
+		Username: user.Username,
+		Setting:  user.Setting,
+		Email:    user.Email,
 	}
 
-	if err := updateUserStatusCache(userId, status == common.UserStatusEnabled); err != nil {
-		return fmt.Errorf("update status cache: %w", err)
+	return userCache, nil
+}
+
+func cacheGetUserBase(userId int) (*UserBase, error) {
+	if !common.RedisEnabled {
+		return nil, fmt.Errorf("redis is not enabled")
 	}
+	var userCache UserBase
+	// Try getting from Redis first
+	err := common.RedisHGetObj(getUserCacheKey(userId), &userCache)
+	if err != nil {
+		return nil, err
+	}
+	return &userCache, nil
+}
 
-	if err := updateUserNameCache(userId, username); err != nil {
-		return fmt.Errorf("update username cache: %w", err)
+// Add atomic quota operations using hash fields
+func cacheIncrUserQuota(userId int, delta int64) error {
+	if !common.RedisEnabled {
+		return nil
 	}
+	return common.RedisHIncrBy(getUserCacheKey(userId), "Quota", delta)
+}
 
-	return nil
+func cacheDecrUserQuota(userId int, delta int64) error {
+	return cacheIncrUserQuota(userId, -delta)
 }
 
-// getUserGroupCache gets user group from cache
+// Helper functions to get individual fields if needed
 func getUserGroupCache(userId int) (string, error) {
-	if !common.RedisEnabled {
-		return "", nil
+	cache, err := GetUserCache(userId)
+	if err != nil {
+		return "", err
 	}
-	return common.RedisGet(fmt.Sprintf(constant.UserGroupKeyFmt, userId))
+	return cache.Group, nil
 }
 
-// getUserQuotaCache gets user quota from cache
 func getUserQuotaCache(userId int) (int, error) {
-	if !common.RedisEnabled {
-		return 0, nil
-	}
-	quotaStr, err := common.RedisGet(fmt.Sprintf(constant.UserQuotaKeyFmt, userId))
+	cache, err := GetUserCache(userId)
 	if err != nil {
 		return 0, err
 	}
-	return strconv.Atoi(quotaStr)
+	return cache.Quota, nil
 }
 
-// getUserStatusCache gets user status from cache
 func getUserStatusCache(userId int) (int, error) {
-	if !common.RedisEnabled {
-		return 0, nil
-	}
-	statusStr, err := common.RedisGet(fmt.Sprintf(constant.UserEnabledKeyFmt, userId))
+	cache, err := GetUserCache(userId)
 	if err != nil {
 		return 0, err
 	}
-	return strconv.Atoi(statusStr)
+	return cache.Status, nil
 }
 
-// getUserNameCache gets username from cache
 func getUserNameCache(userId int) (string, error) {
-	if !common.RedisEnabled {
-		return "", nil
+	cache, err := GetUserCache(userId)
+	if err != nil {
+		return "", err
 	}
-	return common.RedisGet(fmt.Sprintf(constant.UserUsernameKeyFmt, userId))
+	return cache.Username, nil
 }
 
-// getUserCache gets complete user cache
-func getUserCache(userId int) (*userCache, error) {
-	if !common.RedisEnabled {
-		return nil, nil
-	}
-
-	group, err := getUserGroupCache(userId)
+func getUserSettingCache(userId int) (map[string]interface{}, error) {
+	setting := make(map[string]interface{})
+	cache, err := GetUserCache(userId)
 	if err != nil {
-		return nil, fmt.Errorf("get group cache: %w", err)
+		return setting, err
 	}
+	return cache.GetSetting(), nil
+}
 
-	quota, err := getUserQuotaCache(userId)
-	if err != nil {
-		return nil, fmt.Errorf("get quota cache: %w", err)
+// New functions for individual field updates
+func updateUserStatusCache(userId int, status bool) error {
+	if !common.RedisEnabled {
+		return nil
 	}
-
-	status, err := getUserStatusCache(userId)
-	if err != nil {
-		return nil, fmt.Errorf("get status cache: %w", err)
+	statusInt := common.UserStatusEnabled
+	if !status {
+		statusInt = common.UserStatusDisabled
 	}
+	return common.RedisHSetField(getUserCacheKey(userId), "Status", fmt.Sprintf("%d", statusInt))
+}
 
-	username, err := getUserNameCache(userId)
-	if err != nil {
-		return nil, fmt.Errorf("get username cache: %w", err)
+func updateUserQuotaCache(userId int, quota int) error {
+	if !common.RedisEnabled {
+		return nil
 	}
+	return common.RedisHSetField(getUserCacheKey(userId), "Quota", fmt.Sprintf("%d", quota))
+}
 
-	return &userCache{
-		Id:       userId,
-		Group:    group,
-		Quota:    quota,
-		Status:   status,
-		Username: username,
-	}, nil
+func updateUserGroupCache(userId int, group string) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+	return common.RedisHSetField(getUserCacheKey(userId), "Group", group)
 }
 
-// Add atomic quota operations
-func cacheIncrUserQuota(userId int, delta int64) error {
+func updateUserNameCache(userId int, username string) error {
 	if !common.RedisEnabled {
 		return nil
 	}
-	key := fmt.Sprintf(constant.UserQuotaKeyFmt, userId)
-	return common.RedisIncr(key, delta)
+	return common.RedisHSetField(getUserCacheKey(userId), "Username", username)
 }
 
-func cacheDecrUserQuota(userId int, delta int64) error {
-	return cacheIncrUserQuota(userId, -delta)
+func updateUserSettingCache(userId int, setting string) error {
+	if !common.RedisEnabled {
+		return nil
+	}
+	return common.RedisHSetField(getUserCacheKey(userId), "Setting", setting)
 }

relay/channel/cloudflare/adaptor.go

@@ -4,13 +4,14 @@ import (
 	"bytes"
 	"errors"
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
 	"one-api/dto"
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
 	"one-api/relay/constant"
+
+	"github.com/gin-gonic/gin"
 )
 
 type Adaptor struct {

relay/channel/deepseek/adaptor.go

@@ -10,6 +10,7 @@ import (
 	"one-api/relay/channel"
 	"one-api/relay/channel/openai"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/constant"
 )
 
 type Adaptor struct {
@@ -29,7 +30,12 @@ func (a *Adaptor) Init(info *relaycommon.RelayInfo) {
 }
 
 func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
-	return fmt.Sprintf("%s/v1/chat/completions", info.BaseUrl), nil
+	switch info.RelayMode {
+	case constant.RelayModeCompletions:
+		return fmt.Sprintf("%s/beta/completions", info.BaseUrl), nil
+	default:
+		return fmt.Sprintf("%s/v1/chat/completions", info.BaseUrl), nil
+	}
 }
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Header, info *relaycommon.RelayInfo) error {

relay/channel/gemini/adaptor.go

@@ -1,15 +1,21 @@
 package gemini
 
 import (
+	"encoding/json"
 	"errors"
 	"fmt"
-	"github.com/gin-gonic/gin"
 	"io"
 	"net/http"
+	"one-api/common"
 	"one-api/constant"
 	"one-api/dto"
 	"one-api/relay/channel"
 	relaycommon "one-api/relay/common"
+	"one-api/service"
+
+	"strings"
+
+	"github.com/gin-gonic/gin"
 )
 
 type Adaptor struct {
@@ -21,8 +27,36 @@ func (a *Adaptor) ConvertAudioRequest(c *gin.Context, info *relaycommon.RelayInf
 }
 
 func (a *Adaptor) ConvertImageRequest(c *gin.Context, info *relaycommon.RelayInfo, request dto.ImageRequest) (any, error) {
-	//TODO implement me
-	return nil, errors.New("not implemented")
+	if !strings.HasPrefix(info.UpstreamModelName, "imagen") {
+		return nil, errors.New("not supported model for image generation")
+	}
+
+	// convert size to aspect ratio
+	aspectRatio := "1:1" // default aspect ratio
+	switch request.Size {
+	case "1024x1024":
+		aspectRatio = "1:1"
+	case "1024x1792":
+		aspectRatio = "9:16"
+	case "1792x1024":
+		aspectRatio = "16:9"
+	}
+
+	// build gemini imagen request
+	geminiRequest := GeminiImageRequest{
+		Instances: []GeminiImageInstance{
+			{
+				Prompt: request.Prompt,
+			},
+		},
+		Parameters: GeminiImageParameters{
+			SampleCount:      request.N,
+			AspectRatio:      aspectRatio,
+			PersonGeneration: "allow_adult", // default allow adult
+		},
+	}
+
+	return geminiRequest, nil
 }
 
 func (a *Adaptor) Init(info *relaycommon.RelayInfo) {
@@ -40,6 +74,10 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 		}
 	}
 
+	if strings.HasPrefix(info.UpstreamModelName, "imagen") {
+		return fmt.Sprintf("%s/%s/models/%s:predict", info.BaseUrl, version, info.UpstreamModelName), nil
+	}
+
 	action := "generateContent"
 	if info.IsStream {
 		action = "streamGenerateContent?alt=sse"
@@ -73,12 +111,15 @@ func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.Rela
 	return nil, errors.New("not implemented")
 }
 
-
 func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (any, error) {
 	return channel.DoApiRequest(a, c, info, requestBody)
 }
 
 func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+	if strings.HasPrefix(info.UpstreamModelName, "imagen") {
+		return GeminiImageHandler(c, resp, info)
+	}
+
 	if info.IsStream {
 		err, usage = GeminiChatStreamHandler(c, resp, info)
 	} else {
@@ -87,6 +128,60 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 	return
 }
 
+func GeminiImageHandler(c *gin.Context, resp *http.Response, info *relaycommon.RelayInfo) (usage any, err *dto.OpenAIErrorWithStatusCode) {
+	responseBody, readErr := io.ReadAll(resp.Body)
+	if readErr != nil {
+		return nil, service.OpenAIErrorWrapper(readErr, "read_response_body_failed", http.StatusInternalServerError)
+	}
+	_ = resp.Body.Close()
+
+	var geminiResponse GeminiImageResponse
+	if jsonErr := json.Unmarshal(responseBody, &geminiResponse); jsonErr != nil {
+		return nil, service.OpenAIErrorWrapper(jsonErr, "unmarshal_response_body_failed", http.StatusInternalServerError)
+	}
+
+	if len(geminiResponse.Predictions) == 0 {
+		return nil, service.OpenAIErrorWrapper(errors.New("no images generated"), "no_images", http.StatusBadRequest)
+	}
+
+	// convert to openai format response
+	openAIResponse := dto.ImageResponse{
+		Created: common.GetTimestamp(),
+		Data:    make([]dto.ImageData, 0, len(geminiResponse.Predictions)),
+	}
+
+	for _, prediction := range geminiResponse.Predictions {
+		if prediction.RaiFilteredReason != "" {
+			continue // skip filtered image
+		}
+		openAIResponse.Data = append(openAIResponse.Data, dto.ImageData{
+			B64Json: prediction.BytesBase64Encoded,
+		})
+	}
+
+	jsonResponse, jsonErr := json.Marshal(openAIResponse)
+	if jsonErr != nil {
+		return nil, service.OpenAIErrorWrapper(jsonErr, "marshal_response_failed", http.StatusInternalServerError)
+	}
+
+	c.Writer.Header().Set("Content-Type", "application/json")
+	c.Writer.WriteHeader(resp.StatusCode)
+	_, _ = c.Writer.Write(jsonResponse)
+
+	// https://github.com/google-gemini/cookbook/blob/719a27d752aac33f39de18a8d3cb42a70874917e/quickstarts/Counting_Tokens.ipynb
+	// each image has fixed 258 tokens
+	const imageTokens = 258
+	generatedImages := len(openAIResponse.Data)
+
+	usage = &dto.Usage{
+		PromptTokens:     imageTokens * generatedImages, // each generated image has fixed 258 tokens
+		CompletionTokens: 0,                             // image generation does not calculate completion tokens
+		TotalTokens:      imageTokens * generatedImages,
+	}
+
+	return usage, nil
+}
+
 func (a *Adaptor) GetModelList() []string {
 	return ModelList
 }

relay/channel/gemini/constant.go

@@ -16,6 +16,8 @@ var ModelList = []string{
 	"gemini-2.0-pro-exp",
 	// thinking exp
 	"gemini-2.0-flash-thinking-exp",
+	// imagen models
+	"imagen-3.0-generate-002",
 }
 
 var ChannelName = "google gemini"

relay/channel/gemini/dto.go

@@ -109,3 +109,30 @@ type GeminiUsageMetadata struct {
 	CandidatesTokenCount int `json:"candidatesTokenCount"`
 	TotalTokenCount      int `json:"totalTokenCount"`
 }
+
+// Imagen related structs
+type GeminiImageRequest struct {
+	Instances  []GeminiImageInstance `json:"instances"`
+	Parameters GeminiImageParameters `json:"parameters"`
+}
+
+type GeminiImageInstance struct {
+	Prompt string `json:"prompt"`
+}
+
+type GeminiImageParameters struct {
+	SampleCount      int    `json:"sampleCount,omitempty"`
+	AspectRatio      string `json:"aspectRatio,omitempty"`
+	PersonGeneration string `json:"personGeneration,omitempty"`
+}
+
+type GeminiImageResponse struct {
+	Predictions []GeminiImagePrediction `json:"predictions"`
+}
+
+type GeminiImagePrediction struct {
+	MimeType           string `json:"mimeType"`
+	BytesBase64Encoded string `json:"bytesBase64Encoded"`
+	RaiFilteredReason  string `json:"raiFilteredReason,omitempty"`
+	SafetyAttributes   any    `json:"safetyAttributes,omitempty"`
+}

relay/channel/mistral/adaptor.go

@@ -41,9 +41,7 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, info *relaycommon.RelayInfo, re
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-	mistralReq := requestOpenAI2Mistral(*request)
-	//common.LogJson(c, "body", mistralReq)
-	return mistralReq, nil
+	return requestOpenAI2Mistral(request), nil
 }
 
 func (a *Adaptor) ConvertRerankRequest(c *gin.Context, relayMode int, request dto.RerankRequest) (any, error) {
@@ -55,7 +53,6 @@ func (a *Adaptor) ConvertEmbeddingRequest(c *gin.Context, info *relaycommon.Rela
 	return nil, errors.New("not implemented")
 }
 
-
 func (a *Adaptor) DoRequest(c *gin.Context, info *relaycommon.RelayInfo, requestBody io.Reader) (any, error) {
 	return channel.DoApiRequest(a, c, info, requestBody)
 }

relay/channel/mistral/text.go

@@ -1,25 +1,21 @@
 package mistral
 
 import (
-	"encoding/json"
 	"one-api/dto"
 )
 
-func requestOpenAI2Mistral(request dto.GeneralOpenAIRequest) *dto.GeneralOpenAIRequest {
+func requestOpenAI2Mistral(request *dto.GeneralOpenAIRequest) *dto.GeneralOpenAIRequest {
 	messages := make([]dto.Message, 0, len(request.Messages))
 	for _, message := range request.Messages {
-		if !message.IsStringContent() {
-			mediaMessages := message.ParseContent()
-			for j, mediaMessage := range mediaMessages {
-				if mediaMessage.Type == dto.ContentTypeImageURL {
-					imageUrl := mediaMessage.ImageUrl.(dto.MessageImageUrl)
-					mediaMessage.ImageUrl = imageUrl.Url
-					mediaMessages[j] = mediaMessage
-				}
+		mediaMessages := message.ParseContent()
+		for j, mediaMessage := range mediaMessages {
+			if mediaMessage.Type == dto.ContentTypeImageURL {
+				imageUrl := mediaMessage.ImageUrl.(dto.MessageImageUrl)
+				mediaMessage.ImageUrl = imageUrl.Url
+				mediaMessages[j] = mediaMessage
 			}
-			messageRaw, _ := json.Marshal(mediaMessages)
-			message.Content = messageRaw
 		}
+		message.SetMediaContent(mediaMessages)
 		messages = append(messages, dto.Message{
 			Role:       message.Role,
 			Content:    message.Content,

relay/channel/ollama/adaptor.go

@@ -39,6 +39,7 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 
 func (a *Adaptor) SetupRequestHeader(c *gin.Context, req *http.Header, info *relaycommon.RelayInfo) error {
 	channel.SetupApiRequestHeader(info, c, req)
+	req.Set("Authorization", "Bearer "+info.ApiKey)
 	return nil
 }
 
@@ -46,7 +47,7 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, info *relaycommon.RelayInfo, re
 	if request == nil {
 		return nil, errors.New("request is nil")
 	}
-	return requestOpenAI2Ollama(*request), nil
+	return requestOpenAI2Ollama(*request)
 }
 
 func (a *Adaptor) ConvertRerankRequest(c *gin.Context, relayMode int, request dto.RerankRequest) (any, error) {

relay/channel/ollama/dto.go

@@ -3,18 +3,21 @@ package ollama
 import "one-api/dto"
 
 type OllamaRequest struct {
-	Model            string         `json:"model,omitempty"`
-	Messages         []dto.Message  `json:"messages,omitempty"`
-	Stream           bool           `json:"stream,omitempty"`
-	Temperature      *float64       `json:"temperature,omitempty"`
-	Seed             float64        `json:"seed,omitempty"`
-	Topp             float64        `json:"top_p,omitempty"`
-	TopK             int            `json:"top_k,omitempty"`
-	Stop             any            `json:"stop,omitempty"`
-	Tools            []dto.ToolCall `json:"tools,omitempty"`
-	ResponseFormat   any            `json:"response_format,omitempty"`
-	FrequencyPenalty float64        `json:"frequency_penalty,omitempty"`
-	PresencePenalty  float64        `json:"presence_penalty,omitempty"`
+	Model            string             `json:"model,omitempty"`
+	Messages         []dto.Message      `json:"messages,omitempty"`
+	Stream           bool               `json:"stream,omitempty"`
+	Temperature      *float64           `json:"temperature,omitempty"`
+	Seed             float64            `json:"seed,omitempty"`
+	Topp             float64            `json:"top_p,omitempty"`
+	TopK             int                `json:"top_k,omitempty"`
+	Stop             any                `json:"stop,omitempty"`
+	Tools            []dto.ToolCall     `json:"tools,omitempty"`
+	ResponseFormat   any                `json:"response_format,omitempty"`
+	FrequencyPenalty float64            `json:"frequency_penalty,omitempty"`
+	PresencePenalty  float64            `json:"presence_penalty,omitempty"`
+	Suffix           any                `json:"suffix,omitempty"`
+	StreamOptions    *dto.StreamOptions `json:"stream_options,omitempty"`
+	Prompt           any                `json:"prompt,omitempty"`
 }
 
 type Options struct {
@@ -35,7 +38,7 @@ type OllamaEmbeddingRequest struct {
 }
 
 type OllamaEmbeddingResponse struct {
-	Error     string    `json:"error,omitempty"`
-	Model     string    `json:"model"`
+	Error     string      `json:"error,omitempty"`
+	Model     string      `json:"model"`
 	Embedding [][]float64 `json:"embeddings,omitempty"`
 }

relay/channel/ollama/relay-ollama.go

@@ -9,14 +9,36 @@ import (
 	"net/http"
 	"one-api/dto"
 	"one-api/service"
+	"strings"
 )
 
-func requestOpenAI2Ollama(request dto.GeneralOpenAIRequest) *OllamaRequest {
+func requestOpenAI2Ollama(request dto.GeneralOpenAIRequest) (*OllamaRequest, error) {
 	messages := make([]dto.Message, 0, len(request.Messages))
 	for _, message := range request.Messages {
+		if !message.IsStringContent() {
+			mediaMessages := message.ParseContent()
+			for j, mediaMessage := range mediaMessages {
+				if mediaMessage.Type == dto.ContentTypeImageURL {
+					imageUrl := mediaMessage.ImageUrl.(dto.MessageImageUrl)
+					// check if not base64
+					if strings.HasPrefix(imageUrl.Url, "http") {
+						fileData, err := service.GetFileBase64FromUrl(imageUrl.Url)
+						if err != nil {
+							return nil, err
+						}
+						imageUrl.Url = fmt.Sprintf("data:%s;base64,%s", fileData.MimeType, fileData.Base64Data)
+					}
+					mediaMessage.ImageUrl = imageUrl
+					mediaMessages[j] = mediaMessage
+				}
+			}
+			message.SetMediaContent(mediaMessages)
+		}
 		messages = append(messages, dto.Message{
-			Role:    message.Role,
-			Content: message.Content,
+			Role:       message.Role,
+			Content:    message.Content,
+			ToolCalls:  message.ToolCalls,
+			ToolCallId: message.ToolCallId,
 		})
 	}
 	str, ok := request.Stop.(string)
@@ -39,7 +61,10 @@ func requestOpenAI2Ollama(request dto.GeneralOpenAIRequest) *OllamaRequest {
 		ResponseFormat:   request.ResponseFormat,
 		FrequencyPenalty: request.FrequencyPenalty,
 		PresencePenalty:  request.PresencePenalty,
-	}
+		Prompt:           request.Prompt,
+		StreamOptions:    request.StreamOptions,
+		Suffix:           request.Suffix,
+	}, nil
 }
 
 func requestOpenAI2Embeddings(request dto.EmbeddingRequest) *OllamaEmbeddingRequest {

relay/channel/openai/adaptor.go

@@ -119,7 +119,7 @@ func (a *Adaptor) ConvertRequest(c *gin.Context, info *relaycommon.RelayInfo, re
 			request.MaxCompletionTokens = request.MaxTokens
 			request.MaxTokens = 0
 		}
-		if strings.HasPrefix(request.Model, "o3") {
+		if strings.HasPrefix(request.Model, "o3") || strings.HasPrefix(request.Model, "o1") {
 			request.Temperature = nil
 		}
 		if strings.HasSuffix(request.Model, "-high") {

relay/channel/openai/relay-openai.go

@@ -87,6 +87,9 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 			info.SetFirstResponseTime()
 			ticker.Reset(time.Duration(constant.StreamingTimeout) * time.Second)
 			data := scanner.Text()
+			if common.DebugEnabled {
+				println(data)
+			}
 			if len(data) < 6 { // ignore blank line or wrong format
 				continue
 			}
@@ -162,6 +165,7 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 					//}
 					for _, choice := range streamResponse.Choices {
 						responseTextBuilder.WriteString(choice.Delta.GetContentString())
+						responseTextBuilder.WriteString(choice.Delta.GetReasoningContent())
 						if choice.Delta.ToolCalls != nil {
 							if len(choice.Delta.ToolCalls) > toolCount {
 								toolCount = len(choice.Delta.ToolCalls)
@@ -182,6 +186,7 @@ func OaiStreamHandler(c *gin.Context, resp *http.Response, info *relaycommon.Rel
 				//}
 				for _, choice := range streamResponse.Choices {
 					responseTextBuilder.WriteString(choice.Delta.GetContentString())
+					responseTextBuilder.WriteString(choice.Delta.GetReasoningContent())
 					if choice.Delta.ToolCalls != nil {
 						if len(choice.Delta.ToolCalls) > toolCount {
 							toolCount = len(choice.Delta.ToolCalls)
@@ -273,7 +278,7 @@ func OpenaiHandler(c *gin.Context, resp *http.Response, promptTokens int, model
 	if simpleResponse.Usage.TotalTokens == 0 || (simpleResponse.Usage.PromptTokens == 0 && simpleResponse.Usage.CompletionTokens == 0) {
 		completionTokens := 0
 		for _, choice := range simpleResponse.Choices {
-			ctkm, _ := service.CountTextToken(string(choice.Message.Content), model)
+			ctkm, _ := service.CountTextToken(choice.Message.StringContent()+choice.Message.ReasoningContent, model)
 			completionTokens += ctkm
 		}
 		simpleResponse.Usage = dto.Usage{

relay/channel/siliconflow/adaptor.go

@@ -36,6 +36,8 @@ func (a *Adaptor) GetRequestURL(info *relaycommon.RelayInfo) (string, error) {
 		return fmt.Sprintf("%s/v1/embeddings", info.BaseUrl), nil
 	} else if info.RelayMode == constant.RelayModeChatCompletions {
 		return fmt.Sprintf("%s/v1/chat/completions", info.BaseUrl), nil
+	} else if info.RelayMode == constant.RelayModeCompletions {
+		return fmt.Sprintf("%s/v1/completions", info.BaseUrl), nil
 	}
 	return "", errors.New("invalid relay mode")
 }
@@ -72,6 +74,12 @@ func (a *Adaptor) DoResponse(c *gin.Context, resp *http.Response, info *relaycom
 		} else {
 			err, usage = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
 		}
+	case constant.RelayModeCompletions:
+		if info.IsStream {
+			err, usage = openai.OaiStreamHandler(c, resp, info)
+		} else {
+			err, usage = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
+		}
 	case constant.RelayModeEmbeddings:
 		err, usage = openai.OpenaiHandler(c, resp, info.PromptTokens, info.UpstreamModelName)
 	}

relay/channel/zhipu_4v/relay-zhipu_v4.go

@@ -90,8 +90,7 @@ func requestOpenAI2Zhipu(request dto.GeneralOpenAIRequest) *dto.GeneralOpenAIReq
 					mediaMessages[j] = mediaMessage
 				}
 			}
-			messageRaw, _ := json.Marshal(mediaMessages)
-			message.Content = messageRaw
+			message.SetMediaContent(mediaMessages)
 		}
 		messages = append(messages, dto.Message{
 			Role:       message.Role,

relay/common/relay_info.go

@@ -13,24 +13,24 @@ import (
 )
 
 type RelayInfo struct {
-	ChannelType          int
-	ChannelId            int
-	TokenId              int
-	TokenKey             string
-	UserId               int
-	Group                string
-	TokenUnlimited       bool
-	StartTime            time.Time
-	FirstResponseTime    time.Time
-	setFirstResponse     bool
-	ApiType              int
-	IsStream             bool
-	IsPlayground         bool
-	UsePrice             bool
-	RelayMode            int
-	UpstreamModelName    string
-	OriginModelName      string
-	RecodeModelName      string
+	ChannelType       int
+	ChannelId         int
+	TokenId           int
+	TokenKey          string
+	UserId            int
+	Group             string
+	TokenUnlimited    bool
+	StartTime         time.Time
+	FirstResponseTime time.Time
+	setFirstResponse  bool
+	ApiType           int
+	IsStream          bool
+	IsPlayground      bool
+	UsePrice          bool
+	RelayMode         int
+	UpstreamModelName string
+	OriginModelName   string
+	//RecodeModelName      string
 	RequestURLPath       string
 	ApiVersion           string
 	PromptTokens         int
@@ -39,6 +39,7 @@ type RelayInfo struct {
 	BaseUrl              string
 	SupportStreamOptions bool
 	ShouldIncludeUsage   bool
+	IsModelMapped        bool
 	ClientWs             *websocket.Conn
 	TargetWs             *websocket.Conn
 	InputAudioFormat     string
@@ -50,6 +51,18 @@ type RelayInfo struct {
 	ChannelSetting       map[string]interface{}
 }
 
+// 定义支持流式选项的通道类型
+var streamSupportedChannels = map[int]bool{
+	common.ChannelTypeOpenAI:     true,
+	common.ChannelTypeAnthropic:  true,
+	common.ChannelTypeAws:        true,
+	common.ChannelTypeGemini:     true,
+	common.ChannelCloudflare:     true,
+	common.ChannelTypeAzure:      true,
+	common.ChannelTypeVolcEngine: true,
+	common.ChannelTypeOllama:     true,
+}
+
 func GenRelayInfoWs(c *gin.Context, ws *websocket.Conn) *RelayInfo {
 	info := GenRelayInfo(c)
 	info.ClientWs = ws
@@ -89,12 +102,13 @@ func GenRelayInfo(c *gin.Context) *RelayInfo {
 		FirstResponseTime: startTime.Add(-time.Second),
 		OriginModelName:   c.GetString("original_model"),
 		UpstreamModelName: c.GetString("original_model"),
-		RecodeModelName:   c.GetString("recode_model"),
-		ApiType:           apiType,
-		ApiVersion:        c.GetString("api_version"),
-		ApiKey:            strings.TrimPrefix(c.Request.Header.Get("Authorization"), "Bearer "),
-		Organization:      c.GetString("channel_organization"),
-		ChannelSetting:    channelSetting,
+		//RecodeModelName:   c.GetString("original_model"),
+		IsModelMapped:  false,
+		ApiType:        apiType,
+		ApiVersion:     c.GetString("api_version"),
+		ApiKey:         strings.TrimPrefix(c.Request.Header.Get("Authorization"), "Bearer "),
+		Organization:   c.GetString("channel_organization"),
+		ChannelSetting: channelSetting,
 	}
 	if strings.HasPrefix(c.Request.URL.Path, "/pg") {
 		info.IsPlayground = true
@@ -110,9 +124,7 @@ func GenRelayInfo(c *gin.Context) *RelayInfo {
 	if info.ChannelType == common.ChannelTypeVertexAi {
 		info.ApiVersion = c.GetString("region")
 	}
-	if info.ChannelType == common.ChannelTypeOpenAI || info.ChannelType == common.ChannelTypeAnthropic ||
-		info.ChannelType == common.ChannelTypeAws || info.ChannelType == common.ChannelTypeGemini ||
-		info.ChannelType == common.ChannelCloudflare || info.ChannelType == common.ChannelTypeAzure {
+	if streamSupportedChannels[info.ChannelType] {
 		info.SupportStreamOptions = true
 	}
 	return info

relay/helper/model_mapped.go

@@ -0,0 +1,25 @@
+package helper
+
+import (
+	"encoding/json"
+	"fmt"
+	"github.com/gin-gonic/gin"
+	"one-api/relay/common"
+)
+
+func ModelMappedHelper(c *gin.Context, info *common.RelayInfo) error {
+	// map model name
+	modelMapping := c.GetString("model_mapping")
+	if modelMapping != "" && modelMapping != "{}" {
+		modelMap := make(map[string]string)
+		err := json.Unmarshal([]byte(modelMapping), &modelMap)
+		if err != nil {
+			return fmt.Errorf("unmarshal_model_mapping_failed")
+		}
+		if modelMap[info.OriginModelName] != "" {
+			info.UpstreamModelName = modelMap[info.OriginModelName]
+			info.IsModelMapped = true
+		}
+	}
+	return nil
+}

relay/helper/price.go

@@ -0,0 +1,41 @@
+package helper
+
+import (
+	"github.com/gin-gonic/gin"
+	"one-api/common"
+	relaycommon "one-api/relay/common"
+	"one-api/setting"
+)
+
+type PriceData struct {
+	ModelPrice             float64
+	ModelRatio             float64
+	GroupRatio             float64
+	UsePrice               bool
+	ShouldPreConsumedQuota int
+}
+
+func ModelPriceHelper(c *gin.Context, info *relaycommon.RelayInfo, promptTokens int, maxTokens int) PriceData {
+	modelPrice, usePrice := common.GetModelPrice(info.OriginModelName, false)
+	groupRatio := setting.GetGroupRatio(info.Group)
+	var preConsumedQuota int
+	var modelRatio float64
+	if !usePrice {
+		preConsumedTokens := common.PreConsumedQuota
+		if maxTokens != 0 {
+			preConsumedTokens = promptTokens + maxTokens
+		}
+		modelRatio = common.GetModelRatio(info.OriginModelName)
+		ratio := modelRatio * groupRatio
+		preConsumedQuota = int(float64(preConsumedTokens) * ratio)
+	} else {
+		preConsumedQuota = int(modelPrice * common.QuotaPerUnit * groupRatio)
+	}
+	return PriceData{
+		ModelPrice:             modelPrice,
+		ModelRatio:             modelRatio,
+		GroupRatio:             groupRatio,
+		UsePrice:               usePrice,
+		ShouldPreConsumedQuota: preConsumedQuota,
+	}
+}

relay/relay-audio.go

@@ -1,7 +1,6 @@
 package relay
 
 import (
-	"encoding/json"
 	"errors"
 	"fmt"
 	"github.com/gin-gonic/gin"
@@ -11,8 +10,10 @@ import (
 	"one-api/model"
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
+	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting"
+	"strings"
 )
 
 func getAndValidAudioRequest(c *gin.Context, info *relaycommon.RelayInfo) (*dto.AudioRequest, error) {
@@ -27,8 +28,9 @@ func getAndValidAudioRequest(c *gin.Context, info *relaycommon.RelayInfo) (*dto.
 			return nil, errors.New("model is required")
 		}
 		if setting.ShouldCheckPromptSensitive() {
-			err := service.CheckSensitiveInput(audioRequest.Input)
+			words, err := service.CheckSensitiveInput(audioRequest.Input)
 			if err != nil {
+				common.LogWarn(c, fmt.Sprintf("user sensitive words detected: %s", strings.Join(words, ",")))
 				return nil, err
 			}
 		}
@@ -73,15 +75,13 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		relayInfo.PromptTokens = promptTokens
 	}
 
-	modelRatio := common.GetModelRatio(audioRequest.Model)
-	groupRatio := setting.GetGroupRatio(relayInfo.Group)
-	ratio := modelRatio * groupRatio
-	preConsumedQuota := int(float64(preConsumedTokens) * ratio)
+	priceData := helper.ModelPriceHelper(c, relayInfo, preConsumedTokens, 0)
+
 	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
 	if err != nil {
 		return service.OpenAIErrorWrapperLocal(err, "get_user_quota_failed", http.StatusInternalServerError)
 	}
-	preConsumedQuota, userQuota, openaiErr = preConsumeQuota(c, preConsumedQuota, relayInfo)
+	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, priceData.ShouldPreConsumedQuota, relayInfo)
 	if openaiErr != nil {
 		return openaiErr
 	}
@@ -91,19 +91,12 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		}
 	}()
 
-	// map model name
-	modelMapping := c.GetString("model_mapping")
-	if modelMapping != "" {
-		modelMap := make(map[string]string)
-		err := json.Unmarshal([]byte(modelMapping), &modelMap)
-		if err != nil {
-			return service.OpenAIErrorWrapper(err, "unmarshal_model_mapping_failed", http.StatusInternalServerError)
-		}
-		if modelMap[audioRequest.Model] != "" {
-			audioRequest.Model = modelMap[audioRequest.Model]
-		}
+	err = helper.ModelMappedHelper(c, relayInfo)
+	if err != nil {
+		return service.OpenAIErrorWrapperLocal(err, "model_mapped_error", http.StatusInternalServerError)
 	}
-	relayInfo.UpstreamModelName = audioRequest.Model
+
+	audioRequest.Model = relayInfo.UpstreamModelName
 
 	adaptor := GetAdaptor(relayInfo.ApiType)
 	if adaptor == nil {
@@ -140,7 +133,7 @@ func AudioHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		return openaiErr
 	}
 
-	postConsumeQuota(c, relayInfo, audioRequest.Model, usage.(*dto.Usage), ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, 0, false, "")
+	postConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")
 
 	return nil
 }

relay/relay-image.go

@@ -12,6 +12,7 @@ import (
 	"one-api/dto"
 	"one-api/model"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting"
 	"strings"
@@ -60,15 +61,16 @@ func getAndValidImageRequest(c *gin.Context, info *relaycommon.RelayInfo) (*dto.
 	//	return service.OpenAIErrorWrapper(errors.New("n must be between 1 and 10"), "invalid_field_value", http.StatusBadRequest)
 	//}
 	if setting.ShouldCheckPromptSensitive() {
-		err := service.CheckSensitiveInput(imageRequest.Prompt)
+		words, err := service.CheckSensitiveInput(imageRequest.Prompt)
 		if err != nil {
+			common.LogWarn(c, fmt.Sprintf("user sensitive words detected: %s", strings.Join(words, ",")))
 			return nil, err
 		}
 	}
 	return imageRequest, nil
 }
 
-func ImageHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
+func ImageHelper(c *gin.Context) *dto.OpenAIErrorWithStatusCode {
 	relayInfo := relaycommon.GenRelayInfo(c)
 
 	imageRequest, err := getAndValidImageRequest(c, relayInfo)
@@ -77,29 +79,20 @@ func ImageHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 		return service.OpenAIErrorWrapper(err, "invalid_image_request", http.StatusBadRequest)
 	}
 
-	// map model name
-	modelMapping := c.GetString("model_mapping")
-	if modelMapping != "" {
-		modelMap := make(map[string]string)
-		err := json.Unmarshal([]byte(modelMapping), &modelMap)
-		if err != nil {
-			return service.OpenAIErrorWrapper(err, "unmarshal_model_mapping_failed", http.StatusInternalServerError)
-		}
-		if modelMap[imageRequest.Model] != "" {
-			imageRequest.Model = modelMap[imageRequest.Model]
-		}
+	err = helper.ModelMappedHelper(c, relayInfo)
+	if err != nil {
+		return service.OpenAIErrorWrapperLocal(err, "model_mapped_error", http.StatusInternalServerError)
 	}
-	relayInfo.UpstreamModelName = imageRequest.Model
 
-	modelPrice, success := common.GetModelPrice(imageRequest.Model, true)
-	if !success {
-		modelRatio := common.GetModelRatio(imageRequest.Model)
+	imageRequest.Model = relayInfo.UpstreamModelName
+
+	priceData := helper.ModelPriceHelper(c, relayInfo, 0, 0)
+	if !priceData.UsePrice {
 		// modelRatio 16 = modelPrice $0.04
 		// per 1 modelRatio = $0.04 / 16
-		modelPrice = 0.0025 * modelRatio
+		priceData.ModelPrice = 0.0025 * priceData.ModelRatio
 	}
 
-	groupRatio := setting.GetGroupRatio(relayInfo.Group)
 	userQuota, err := model.GetUserQuota(relayInfo.UserId, false)
 
 	sizeRatio := 1.0
@@ -122,11 +115,11 @@ func ImageHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 		}
 	}
 
-	imageRatio := modelPrice * sizeRatio * qualityRatio * float64(imageRequest.N)
-	quota := int(imageRatio * groupRatio * common.QuotaPerUnit)
+	priceData.ModelPrice *= sizeRatio * qualityRatio * float64(imageRequest.N)
+	quota := int(priceData.ModelPrice * priceData.GroupRatio * common.QuotaPerUnit)
 
 	if userQuota-quota < 0 {
-		return service.OpenAIErrorWrapperLocal(errors.New(fmt.Sprintf("image pre-consumed quota failed, user quota: %d, need quota: %d", userQuota, quota)), "insufficient_user_quota", http.StatusBadRequest)
+		return service.OpenAIErrorWrapperLocal(fmt.Errorf("image pre-consumed quota failed, user quota: %s, need quota: %s", common.FormatQuota(userQuota), common.FormatQuota(quota)), "insufficient_user_quota", http.StatusForbidden)
 	}
 
 	adaptor := GetAdaptor(relayInfo.ApiType)
@@ -184,7 +177,6 @@ func ImageHelper(c *gin.Context, relayMode int) *dto.OpenAIErrorWithStatusCode {
 	}
 
 	logContent := fmt.Sprintf("大小 %s, 品质 %s", imageRequest.Size, quality)
-	postConsumeQuota(c, relayInfo, imageRequest.Model, usage, 0, 0, userQuota, 0, groupRatio, imageRatio, true, logContent)
-
+	postConsumeQuota(c, relayInfo, usage, 0, userQuota, priceData, logContent)
 	return nil
 }

relay/relay-mj.go

@@ -194,7 +194,7 @@ func RelaySwapFace(c *gin.Context) *dto.MidjourneyResponse {
 	}
 	defer func(ctx context.Context) {
 		if mjResp.StatusCode == 200 && mjResp.Response.Code == 1 {
-			err := model.PostConsumeQuota(relayInfo, userQuota, quota, 0, true)
+			err := service.PostConsumeQuota(relayInfo, quota, 0, true)
 			if err != nil {
 				common.SysError("error consuming token remain quota: " + err.Error())
 			}
@@ -500,7 +500,7 @@ func RelayMidjourneySubmit(c *gin.Context, relayMode int) *dto.MidjourneyRespons
 
 	defer func(ctx context.Context) {
 		if consumeQuota && midjResponseWithStatus.StatusCode == 200 {
-			err := model.PostConsumeQuota(relayInfo, userQuota, quota, 0, true)
+			err := service.PostConsumeQuota(relayInfo, quota, 0, true)
 			if err != nil {
 				common.SysError("error consuming token remain quota: " + err.Error())
 			}

relay/relay-text.go

@@ -5,6 +5,7 @@ import (
 	"encoding/json"
 	"errors"
 	"fmt"
+	"github.com/bytedance/gopkg/util/gopool"
 	"io"
 	"math"
 	"net/http"
@@ -14,6 +15,7 @@ import (
 	"one-api/model"
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
+	"one-api/relay/helper"
 	"one-api/service"
 	"one-api/setting"
 	"strings"
@@ -75,40 +77,21 @@ func TextHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		return service.OpenAIErrorWrapperLocal(err, "invalid_text_request", http.StatusBadRequest)
 	}
 
-	// map model name
-	//isModelMapped := false
-	modelMapping := c.GetString("model_mapping")
-	//isModelMapped := false
-	if modelMapping != "" && modelMapping != "{}" {
-		modelMap := make(map[string]string)
-		err := json.Unmarshal([]byte(modelMapping), &modelMap)
-		if err != nil {
-			return service.OpenAIErrorWrapperLocal(err, "unmarshal_model_mapping_failed", http.StatusInternalServerError)
-		}
-		if modelMap[textRequest.Model] != "" {
-			//isModelMapped = true
-			textRequest.Model = modelMap[textRequest.Model]
-			// set upstream model name
-			//isModelMapped = true
-		}
-	}
-	relayInfo.UpstreamModelName = textRequest.Model
-	relayInfo.RecodeModelName = textRequest.Model
-	modelPrice, getModelPriceSuccess := common.GetModelPrice(textRequest.Model, false)
-	groupRatio := setting.GetGroupRatio(relayInfo.Group)
-
-	var preConsumedQuota int
-	var ratio float64
-	var modelRatio float64
-	//err := service.SensitiveWordsCheck(textRequest)
-
 	if setting.ShouldCheckPromptSensitive() {
-		err = checkRequestSensitive(textRequest, relayInfo)
+		words, err := checkRequestSensitive(textRequest, relayInfo)
 		if err != nil {
+			common.LogWarn(c, fmt.Sprintf("user sensitive words detected: %s", strings.Join(words, ", ")))
 			return service.OpenAIErrorWrapperLocal(err, "sensitive_words_detected", http.StatusBadRequest)
 		}
 	}
 
+	err = helper.ModelMappedHelper(c, relayInfo)
+	if err != nil {
+		return service.OpenAIErrorWrapperLocal(err, "model_mapped_error", http.StatusInternalServerError)
+	}
+
+	textRequest.Model = relayInfo.UpstreamModelName
+
 	// 获取 promptTokens，如果上下文中已经存在，则直接使用
 	var promptTokens int
 	if value, exists := c.Get("prompt_tokens"); exists {
@@ -123,20 +106,10 @@ func TextHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		c.Set("prompt_tokens", promptTokens)
 	}
 
-	if !getModelPriceSuccess {
-		preConsumedTokens := common.PreConsumedQuota
-		if textRequest.MaxTokens != 0 {
-			preConsumedTokens = promptTokens + int(textRequest.MaxTokens)
-		}
-		modelRatio = common.GetModelRatio(textRequest.Model)
-		ratio = modelRatio * groupRatio
-		preConsumedQuota = int(float64(preConsumedTokens) * ratio)
-	} else {
-		preConsumedQuota = int(modelPrice * common.QuotaPerUnit * groupRatio)
-	}
+	priceData := helper.ModelPriceHelper(c, relayInfo, promptTokens, int(textRequest.MaxTokens))
 
 	// pre-consume quota 预消耗配额
-	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, preConsumedQuota, relayInfo)
+	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, priceData.ShouldPreConsumedQuota, relayInfo)
 	if openaiErr != nil {
 		return openaiErr
 	}
@@ -219,10 +192,10 @@ func TextHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode) {
 		return openaiErr
 	}
 
-	if strings.HasPrefix(relayInfo.RecodeModelName, "gpt-4o-audio") {
-		service.PostAudioConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, getModelPriceSuccess, "")
+	if strings.HasPrefix(relayInfo.OriginModelName, "gpt-4o-audio") {
+		service.PostAudioConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")
 	} else {
-		postConsumeQuota(c, relayInfo, relayInfo.RecodeModelName, usage.(*dto.Usage), ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, getModelPriceSuccess, "")
+		postConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")
 	}
 	return nil
 }
@@ -247,19 +220,20 @@ func getPromptTokens(textRequest *dto.GeneralOpenAIRequest, info *relaycommon.Re
 	return promptTokens, err
 }
 
-func checkRequestSensitive(textRequest *dto.GeneralOpenAIRequest, info *relaycommon.RelayInfo) error {
+func checkRequestSensitive(textRequest *dto.GeneralOpenAIRequest, info *relaycommon.RelayInfo) ([]string, error) {
 	var err error
+	var words []string
 	switch info.RelayMode {
 	case relayconstant.RelayModeChatCompletions:
-		err = service.CheckSensitiveMessages(textRequest.Messages)
+		words, err = service.CheckSensitiveMessages(textRequest.Messages)
 	case relayconstant.RelayModeCompletions:
-		err = service.CheckSensitiveInput(textRequest.Prompt)
+		words, err = service.CheckSensitiveInput(textRequest.Prompt)
 	case relayconstant.RelayModeModerations:
-		err = service.CheckSensitiveInput(textRequest.Input)
+		words, err = service.CheckSensitiveInput(textRequest.Input)
 	case relayconstant.RelayModeEmbeddings:
-		err = service.CheckSensitiveInput(textRequest.Input)
+		words, err = service.CheckSensitiveInput(textRequest.Input)
 	}
-	return err
+	return words, err
 }
 
 // 预扣费并返回用户剩余配额
@@ -272,7 +246,7 @@ func preConsumeQuota(c *gin.Context, preConsumedQuota int, relayInfo *relaycommo
 		return 0, 0, service.OpenAIErrorWrapperLocal(errors.New("user quota is not enough"), "insufficient_user_quota", http.StatusForbidden)
 	}
 	if userQuota-preConsumedQuota < 0 {
-		return 0, 0, service.OpenAIErrorWrapperLocal(fmt.Errorf("chat pre-consumed quota failed, user quota: %d, need quota: %d", userQuota, preConsumedQuota), "insufficient_user_quota", http.StatusBadRequest)
+		return 0, 0, service.OpenAIErrorWrapperLocal(fmt.Errorf("chat pre-consumed quota failed, user quota: %s, need quota: %s", common.FormatQuota(userQuota), common.FormatQuota(preConsumedQuota)), "insufficient_user_quota", http.StatusForbidden)
 	}
 	if userQuota > 100*preConsumedQuota {
 		// 用户额度充足，判断令牌额度是否充足
@@ -282,18 +256,18 @@ func preConsumeQuota(c *gin.Context, preConsumedQuota int, relayInfo *relaycommo
 			if tokenQuota > 100*preConsumedQuota {
 				// 令牌额度充足，信任令牌
 				preConsumedQuota = 0
-				common.LogInfo(c, fmt.Sprintf("user %d quota %d and token %d quota %d are enough, trusted and no need to pre-consume", relayInfo.UserId, userQuota, relayInfo.TokenId, tokenQuota))
+				common.LogInfo(c, fmt.Sprintf("user %d quota %s and token %d quota %d are enough, trusted and no need to pre-consume", relayInfo.UserId, common.FormatQuota(userQuota), relayInfo.TokenId, tokenQuota))
 			}
 		} else {
 			// in this case, we do not pre-consume quota
 			// because the user has enough quota
 			preConsumedQuota = 0
-			common.LogInfo(c, fmt.Sprintf("user %d with unlimited token has enough quota %d, trusted and no need to pre-consume", relayInfo.UserId, userQuota))
+			common.LogInfo(c, fmt.Sprintf("user %d with unlimited token has enough quota %s, trusted and no need to pre-consume", relayInfo.UserId, common.FormatQuota(userQuota)))
 		}
 	}
 
 	if preConsumedQuota > 0 {
-		err = model.PreConsumeTokenQuota(relayInfo, preConsumedQuota)
+		err = service.PreConsumeTokenQuota(relayInfo, preConsumedQuota)
 		if err != nil {
 			return 0, 0, service.OpenAIErrorWrapperLocal(err, "pre_consume_token_quota_failed", http.StatusForbidden)
 		}
@@ -307,20 +281,19 @@ func preConsumeQuota(c *gin.Context, preConsumedQuota int, relayInfo *relaycommo
 
 func returnPreConsumedQuota(c *gin.Context, relayInfo *relaycommon.RelayInfo, userQuota int, preConsumedQuota int) {
 	if preConsumedQuota != 0 {
-		go func() {
+		gopool.Go(func() {
 			relayInfoCopy := *relayInfo
 
-			err := model.PostConsumeQuota(&relayInfoCopy, userQuota, -preConsumedQuota, 0, false)
+			err := service.PostConsumeQuota(&relayInfoCopy, -preConsumedQuota, 0, false)
 			if err != nil {
 				common.SysError("error return pre-consumed quota: " + err.Error())
 			}
-		}()
+		})
 	}
 }
 
-func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, modelName string,
-	usage *dto.Usage, ratio float64, preConsumedQuota int, userQuota int, modelRatio float64, groupRatio float64,
-	modelPrice float64, usePrice bool, extraContent string) {
+func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo,
+	usage *dto.Usage, preConsumedQuota int, userQuota int, priceData helper.PriceData, extraContent string) {
 	if usage == nil {
 		usage = &dto.Usage{
 			PromptTokens:     relayInfo.PromptTokens,
@@ -332,12 +305,18 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, modelN
 	useTimeSeconds := time.Now().Unix() - relayInfo.StartTime.Unix()
 	promptTokens := usage.PromptTokens
 	completionTokens := usage.CompletionTokens
+	modelName := relayInfo.OriginModelName
 
 	tokenName := ctx.GetString("token_name")
 	completionRatio := common.GetCompletionRatio(modelName)
+	ratio := priceData.ModelRatio * priceData.GroupRatio
+	modelRatio := priceData.ModelRatio
+	groupRatio := priceData.GroupRatio
+	modelPrice := priceData.ModelPrice
+	usePrice := priceData.UsePrice
 
 	quota := 0
-	if !usePrice {
+	if !priceData.UsePrice {
 		quota = promptTokens + int(math.Round(float64(completionTokens)*completionRatio))
 		quota = int(math.Round(float64(quota) * ratio))
 		if ratio != 0 && quota <= 0 {
@@ -368,7 +347,7 @@ func postConsumeQuota(ctx *gin.Context, relayInfo *relaycommon.RelayInfo, modelN
 		//}
 		quotaDelta := quota - preConsumedQuota
 		if quotaDelta != 0 {
-			err := model.PostConsumeQuota(relayInfo, userQuota, quotaDelta, preConsumedQuota, true)
+			err := service.PostConsumeQuota(relayInfo, quotaDelta, preConsumedQuota, true)
 			if err != nil {
 				common.LogError(ctx, "error consuming token remain quota: "+err.Error())
 			}

relay/relay_embedding.go

@@ -10,8 +10,8 @@ import (
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
 	relayconstant "one-api/relay/constant"
+	"one-api/relay/helper"
 	"one-api/service"
-	"one-api/setting"
 )
 
 func getEmbeddingPromptToken(embeddingRequest dto.EmbeddingRequest) int {
@@ -47,43 +47,20 @@ func EmbeddingHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode)
 		return service.OpenAIErrorWrapperLocal(err, "invalid_embedding_request", http.StatusBadRequest)
 	}
 
-	// map model name
-	modelMapping := c.GetString("model_mapping")
-	//isModelMapped := false
-	if modelMapping != "" && modelMapping != "{}" {
-		modelMap := make(map[string]string)
-		err := json.Unmarshal([]byte(modelMapping), &modelMap)
-		if err != nil {
-			return service.OpenAIErrorWrapperLocal(err, "unmarshal_model_mapping_failed", http.StatusInternalServerError)
-		}
-		if modelMap[embeddingRequest.Model] != "" {
-			embeddingRequest.Model = modelMap[embeddingRequest.Model]
-			// set upstream model name
-			//isModelMapped = true
-		}
+	err = helper.ModelMappedHelper(c, relayInfo)
+	if err != nil {
+		return service.OpenAIErrorWrapperLocal(err, "model_mapped_error", http.StatusInternalServerError)
 	}
 
-	relayInfo.UpstreamModelName = embeddingRequest.Model
-	modelPrice, success := common.GetModelPrice(embeddingRequest.Model, false)
-	groupRatio := setting.GetGroupRatio(relayInfo.Group)
-
-	var preConsumedQuota int
-	var ratio float64
-	var modelRatio float64
+	embeddingRequest.Model = relayInfo.UpstreamModelName
 
 	promptToken := getEmbeddingPromptToken(*embeddingRequest)
-	if !success {
-		preConsumedTokens := promptToken
-		modelRatio = common.GetModelRatio(embeddingRequest.Model)
-		ratio = modelRatio * groupRatio
-		preConsumedQuota = int(float64(preConsumedTokens) * ratio)
-	} else {
-		preConsumedQuota = int(modelPrice * common.QuotaPerUnit * groupRatio)
-	}
 	relayInfo.PromptTokens = promptToken
 
+	priceData := helper.ModelPriceHelper(c, relayInfo, promptToken, 0)
+
 	// pre-consume quota 预消耗配额
-	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, preConsumedQuota, relayInfo)
+	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, priceData.ShouldPreConsumedQuota, relayInfo)
 	if openaiErr != nil {
 		return openaiErr
 	}
@@ -132,6 +109,6 @@ func EmbeddingHelper(c *gin.Context) (openaiErr *dto.OpenAIErrorWithStatusCode)
 		service.ResetStatusCode(openaiErr, statusCodeMappingStr)
 		return openaiErr
 	}
-	postConsumeQuota(c, relayInfo, embeddingRequest.Model, usage.(*dto.Usage), ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, success, "")
+	postConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")
 	return nil
 }

relay/relay_rerank.go

@@ -9,8 +9,8 @@ import (
 	"one-api/common"
 	"one-api/dto"
 	relaycommon "one-api/relay/common"
+	"one-api/relay/helper"
 	"one-api/service"
-	"one-api/setting"
 )
 
 func getRerankPromptToken(rerankRequest dto.RerankRequest) int {
@@ -40,43 +40,20 @@ func RerankHelper(c *gin.Context, relayMode int) (openaiErr *dto.OpenAIErrorWith
 		return service.OpenAIErrorWrapperLocal(fmt.Errorf("documents is empty"), "invalid_documents", http.StatusBadRequest)
 	}
 
-	// map model name
-	modelMapping := c.GetString("model_mapping")
-	//isModelMapped := false
-	if modelMapping != "" && modelMapping != "{}" {
-		modelMap := make(map[string]string)
-		err := json.Unmarshal([]byte(modelMapping), &modelMap)
-		if err != nil {
-			return service.OpenAIErrorWrapperLocal(err, "unmarshal_model_mapping_failed", http.StatusInternalServerError)
-		}
-		if modelMap[rerankRequest.Model] != "" {
-			rerankRequest.Model = modelMap[rerankRequest.Model]
-			// set upstream model name
-			//isModelMapped = true
-		}
+	err = helper.ModelMappedHelper(c, relayInfo)
+	if err != nil {
+		return service.OpenAIErrorWrapperLocal(err, "model_mapped_error", http.StatusInternalServerError)
 	}
 
-	relayInfo.UpstreamModelName = rerankRequest.Model
-	modelPrice, success := common.GetModelPrice(rerankRequest.Model, false)
-	groupRatio := setting.GetGroupRatio(relayInfo.Group)
-
-	var preConsumedQuota int
-	var ratio float64
-	var modelRatio float64
+	rerankRequest.Model = relayInfo.UpstreamModelName
 
 	promptToken := getRerankPromptToken(*rerankRequest)
-	if !success {
-		preConsumedTokens := promptToken
-		modelRatio = common.GetModelRatio(rerankRequest.Model)
-		ratio = modelRatio * groupRatio
-		preConsumedQuota = int(float64(preConsumedTokens) * ratio)
-	} else {
-		preConsumedQuota = int(modelPrice * common.QuotaPerUnit * groupRatio)
-	}
 	relayInfo.PromptTokens = promptToken
 
+	priceData := helper.ModelPriceHelper(c, relayInfo, promptToken, 0)
+
 	// pre-consume quota 预消耗配额
-	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, preConsumedQuota, relayInfo)
+	preConsumedQuota, userQuota, openaiErr := preConsumeQuota(c, priceData.ShouldPreConsumedQuota, relayInfo)
 	if openaiErr != nil {
 		return openaiErr
 	}
@@ -124,6 +101,6 @@ func RerankHelper(c *gin.Context, relayMode int) (openaiErr *dto.OpenAIErrorWith
 		service.ResetStatusCode(openaiErr, statusCodeMappingStr)
 		return openaiErr
 	}
-	postConsumeQuota(c, relayInfo, rerankRequest.Model, usage.(*dto.Usage), ratio, preConsumedQuota, userQuota, modelRatio, groupRatio, modelPrice, success, "")
+	postConsumeQuota(c, relayInfo, usage.(*dto.Usage), preConsumedQuota, userQuota, priceData, "")
 	return nil
 }

relay/relay_task.go

@@ -113,7 +113,7 @@ func RelayTaskSubmit(c *gin.Context, relayMode int) (taskErr *dto.TaskError) {
 		// release quota
 		if relayInfo.ConsumeQuota && taskErr == nil {
 
-			err := model.PostConsumeQuota(relayInfo.ToRelayInfo(), userQuota, quota, 0, true)
+			err := service.PostConsumeQuota(relayInfo.ToRelayInfo(), quota, 0, true)
 			if err != nil {
 				common.SysError("error consuming token remain quota: " + err.Error())
 			}

router/api-router.go

@@ -56,6 +56,7 @@ func SetApiRouter(router *gin.Engine) {
 				selfRoute.POST("/pay", controller.RequestEpay)
 				selfRoute.POST("/amount", controller.RequestAmount)
 				selfRoute.POST("/aff_transfer", controller.TransferAffQuota)
+				selfRoute.PUT("/setting", controller.UpdateUserSetting)
 			}
 
 			adminRoute := userRoute.Group("/")